In this episode, host Frank La Vigne and co-host Candice Gillhoolley sit down with Danny Wall, the founder, CEO, and CTO of OA Quantum Labs, for an in-depth conversation about the real-world intersection of quantum computing and artificial intelligence.
You’ll hear Danny Wall pull the curtain back on how OA Quantum Labs is pushing quantum solutions beyond the research phase and into commercially viable applications. From accelerating AI training and inference to spinning out novel materials at lightning speed, Danny shares firsthand stories about quantum-enhanced breakthroughs in material science, finance, and more.
This episode dives into common misconceptions—like the idea that AI is actually running on quantum computers—and Danny explains the nuanced, current reality: quantum as an incredible mathematical accelerator and enhancement for AI, rather than a full replacement. You’ll also get practical advice for developers, researchers, and investors eager to get started with quantum, and insights on what it really takes to stay ahead in a field moving as fast as quantum.
If you’re curious about how quantum technologies are escaping the confines of the lab and making real commercial impact, this is the episode you’ve been waiting for!
Time Stamps
00:00 “Quantum Labs Driving AI Innovation”
03:31 “Quantum Computing Enhances AI Efficiency”
09:32 Advanced Materials Breakthroughs Revolutionizing Industries
12:58 Quantum Investing: Beyond PhD Pedigrees
15:47 “Quantum, Solutions, and Strategic Investment”
18:05 “Jump Into Quantum Development”
20:27 “Quantum Enhancement for AI Solutions”
25:38 AI Limits and Misconceptions
27:02 “AI Creativity Hack with Roles”
33:16 “Challenges in Quantum Error Correction”
36:37 Quantum Computing’s Material Challenges
38:02 “AI Progress Hitting Limits”
42:49 “Quantum Encryption and Neural Networks”
47:19 “Schrödinger’s Cat Explained Simply”
48:16 “Quantum Physics Misconceptions Explained”
Transcript
If we can advance quantum a little bit faster, while
Speaker:quantum comes with a power requirement in the terms
Speaker:of cooling, the actual cost to run the
Speaker:QPU is almost zero, right? It
Speaker:really doesn't cost a whole lot to run a QPU.
Speaker:AI may be approaching its limits, but quantum
Speaker:computing could be the next leap forward.
Speaker:Hello and welcome back to Impact Quantum, the podcast where we explore the emerging
Speaker:field of quantum computing. And you don't need to be a PhD, you just
Speaker:need to be a little bit curious. And with me is the most quantum curious
Speaker:person I know, Candace Cahouli. How's it going, Candace? It's great.
Speaker:Today's a wonderful day. I'm really, really excited.
Speaker:We are going to be speaking with Danny Wall, who
Speaker:is the founder, CEO, and CTO
Speaker:at OA Quantum Labs. Hi,
Speaker:Danny. How are you today? I am fantastic.
Speaker:How about yourself? Doing all right. It's always
Speaker:good to hear from folks in a state warmer and
Speaker:sunnier than where I am. We had our first winter storm warning
Speaker:here for the season here in the Baltimore, D.C. area.
Speaker:And kids, kids were— had a late start to school and that always
Speaker:throws things off. But Candace is an old hat at
Speaker:snow. It's Montreal. They probably already had like 20 feet already for the season.
Speaker:No, seriously, like it's true because it always starts. It
Speaker:usually always starts Halloween. Like you get a little bit in Halloween
Speaker:just to have a taste. So if you're— if your costume does not fit
Speaker:over your winter coat, It is not an acceptable costume
Speaker:here in Montreal, Quebec. But yeah, it's snowing every day.
Speaker:Like, it just snows every day. But that's just how it is. But you
Speaker:learn how to deal with it. And so it's just fine. Just very pretty. We
Speaker:actually get our first snow overnight tonight. Oh,
Speaker:nice. Oh, you must be in the altitude then. About
Speaker:5,000, a little over 5,000 feet. Yeah. Oh, okay. Okay.
Speaker:You're coming to us from sunny New Mexico, or normally sunny New Mexico.
Speaker:And so tell us, what are you doing? We, in the virtual
Speaker:green room, we spoke briefly, working on building something really cool.
Speaker:Yeah, so I'm building a quantum lab
Speaker:out here. So OA
Speaker:Quantum Labs is not just a quantum lab, so we don't do
Speaker:just research. All of our research is
Speaker:100% geared towards
Speaker:creating true commercial application of quantum
Speaker:technology. So a good example is we are also
Speaker:the owners of multiple AI companies,
Speaker:which we have now acquired. So as
Speaker:part of that, we are applying
Speaker:quantum computing in its current state of the
Speaker:science to multiple different
Speaker:components within the AI
Speaker:ecosystem. Interesting. Okay,
Speaker:okay. How so? Like what particularly, like, I'm curious
Speaker:to see what the intersection of quantum and AI, sorry. Okay, so the very first
Speaker:things that we did was reducing
Speaker:training cost and time. That was the
Speaker:easiest place where quantum could make the
Speaker:biggest impact. And this was back when we
Speaker:were still on, you know, sub-100
Speaker:qubit systems, really in the 50s somewhere, logical qubits.
Speaker:Now what we are doing is we're also improving
Speaker:AI inference in a number of areas.
Speaker:So if you, I'm going to
Speaker:oversimplify this a little bit to the point of it almost
Speaker:being wrong, but it provides a good analogy.
Speaker:One of the things that quantum computing is really, really, really
Speaker:good at is math, right? It does
Speaker:math and complex math very, very quickly. So
Speaker:if you think of a quantum computer almost
Speaker:like a super ridiculous
Speaker:calculator, you can use AI
Speaker:for all of its inference, but when math needs to take place,
Speaker:you throw the math to the quantum computer, get the math back, and
Speaker:done, and then it comes back. Where this works the best is
Speaker:in materials. When you're, when you're doing anything with
Speaker:materials or molecules.
Speaker:Interesting. I mean, that
Speaker:makes sense, right? Because there's definitely a tight correlation between
Speaker:quantum effects and chemistry. Yes. And it sounds a
Speaker:bit like, like a GPU, right? In a sense, right? Like almost. You—
Speaker:that's what I said, a bit like, right? Like you're sending off whether it's a
Speaker:video game, whether it's AI or neural network training,
Speaker:you're just saying, here's a bunch of stuff, GPU, go for it, right?
Speaker:Yes. And then you come back with an answer. Yeah, yeah. Only you're saying
Speaker:QPU, go for it. QPU, yeah, right. Yeah,
Speaker:right. I'm hoping that term catches. I'm hoping that term catches on. Yeah, yeah,
Speaker:exactly. We do. We hear it a lot. Yeah. So QPU
Speaker:is actually already a term on that thing, but
Speaker:Quantum computing, the architecture of quantum computers is different.
Speaker:It's not really the same architecture where you have
Speaker:a central processing unit and then memory
Speaker:sits somewhere else, and then you have
Speaker:buses between your CPU and your— it's not like
Speaker:that in the quantum world. Interesting. The
Speaker:memory is, let's call it, on-chip.
Speaker:Right. Well, there's also kind of— there's also the thing, like,
Speaker:once you read the memory, do you collapse the quantum state? I know that once
Speaker:you get into kind of the brass tacks of, you know, beyond
Speaker:like the theoretical, like you start to get some— it starts to get weird
Speaker:real fast, right? Because like, you know, how do you— debugging a
Speaker:quantum system, right? We've already talked with some other guests about that. Like, that's, you
Speaker:know, how do you, you know, if you— how do you step through the code,
Speaker:right? And like you peek at the variables. Well, as soon as you do that
Speaker:in a quantum system, You're collapsing. You're collapsing and
Speaker:you kind of lose the advantage of quantum. Yeah, yeah, yeah. So like, I, I
Speaker:would imagine that there's a lot of these little gotchas that nobody's really fully kind
Speaker:of worked through just yet. Um, so there are, and this
Speaker:is why, um, as long
Speaker:as you understand what the limitations are,
Speaker:quantum has some really significant advantages
Speaker:right now. This is the reason
Speaker:why JP Morgan, as an example, is spending $1.5 billion
Speaker:on quantum computing, because there are certain things, certain
Speaker:mathematics, QAOA, right? Quantum Approximate
Speaker:Optimization Algorithms, right? Where you're using
Speaker:quantum to do
Speaker:certain mathematical functions that just take too
Speaker:long, and they're, you know, take a second or two on quantum,
Speaker:they take minutes on classical. And, and when
Speaker:you're in the world of finance, you know, a minute is too
Speaker:long. Same goes with,
Speaker:um, um, advanced correlation algorithms. You get into quantum advanced
Speaker:correlation algorithms, and those run really
Speaker:ridiculously a lot faster. Right.
Speaker:But not for all problems, just certain. Yes, that's what I'm saying. Yeah.
Speaker:When you, when you understand what problem domains quantum is
Speaker:really good at, it becomes a lot easier,
Speaker:faster to start applying commercial application
Speaker:to it. Gotcha. So you
Speaker:talked about the financial sector. What other, what
Speaker:other industries are you think, do you think are the most primed to benefit
Speaker:first? Okay, so where
Speaker:it's already benefiting is anything where you need molecular
Speaker:or quantum knowledge or effects or whatever, right?
Speaker:So material sciences is a big one.
Speaker:In partnership with Ursulaing Quantum
Speaker:Innovations, we have created the single
Speaker:most advanced materials, let's call
Speaker:it, engineering platform in the
Speaker:world, right? Our nearest
Speaker:competitor is— oh my gosh, I was just going to
Speaker:say them.
Speaker:They just got this massive amount of money and I totally
Speaker:spaced their name. Cusp AI, I think that's what it is.
Speaker:They— so they're supposed to be a material science platform. They need 6
Speaker:months and an entire team of material sciences
Speaker:scientists to do almost anything. And we
Speaker:were spinning out new materials at the pace of a new one every
Speaker:2 weeks. Oh, wow. Okay. Yeah. Like,
Speaker:we have— we got— we created a material that
Speaker:is stronger and harder than
Speaker:carbon fiber, but about half the price to manufacture.
Speaker:We created brand new heat shielding that
Speaker:survives multiple multiple reentries and is
Speaker:far less expensive to produce than what SpaceX is using today.
Speaker:We created a new material, a new
Speaker:advanced material for
Speaker:heat management. It basically pulls heat away to use as like heat sinks
Speaker:and those kinds of things that is far better than
Speaker:anything that exists. So we finally— we were creating so
Speaker:many new materials so fast that we overran the sales team's ability
Speaker:to keep up, so we spun that out into a brand new company,
Speaker:and now that, that guy is off to the races.
Speaker:Um, and, uh, so the other place where,
Speaker:um, it helps a lot is again in modeling, uh, quantum
Speaker:effects. I was able to create a whole brand new
Speaker:GPU kernel that is far better than Flash
Speaker:Attention V2 because I modeled how
Speaker:electrons flow through a GPU and therefore was
Speaker:able to optimize the code for how the attention
Speaker:mechanisms work on inference.
Speaker:Interesting.
Speaker:Interesting.
Speaker:What sorts of hardware does this run on? I'm sorry,
Speaker:Candace. No, no, go ahead. What sorts of hardware? Is it hardware agnostic? Oh, no,
Speaker:no, no. So I mean, I wrote it to be very specific
Speaker:to the NVIDIA H100,
Speaker:A100, and above better, right? That makes sense. Yeah, this
Speaker:is, this, I, when I wrote it, it was when the,
Speaker:uh, uh, X was coming out with all the news about their brand
Speaker:new Colossus supercluster, blah, blah, blah. And I was like, I wonder
Speaker:if I could, since I can model, um, molecules
Speaker:and all that kind of, and quantum effects and all that other kinds of
Speaker:good stuff, can I model how things flow
Speaker:through a GPU and therefore improve on
Speaker:improve on how the attention mechanism
Speaker:works within a GPU, and it's better by a
Speaker:lot. Interesting. Between 1.5 and
Speaker:3x improved inference. Oh, wow.
Speaker:Yeah. Depending on where you
Speaker:are in the stack, do you need sparse attention or—
Speaker:all of a sudden I drew a blank on the name— sparse attention or
Speaker:Wow. I deal with this every day and all of a sudden I blanked. It
Speaker:happens to the best of us. Of course, attention is the thing that they— that
Speaker:you really kind of need the least of.
Speaker:But yeah, so I got you. Okay. I mean, that makes sense.
Speaker:Look, honestly, everything you're talking about is so incredibly exciting. So how
Speaker:do newcomers interested in quantum and AI
Speaker:researchers, entrepreneurs, investors,
Speaker:What advice or first steps would you recommend today to get
Speaker:them involved meaningfully? Okay, so
Speaker:that's really gonna depend on which one of those you're talking
Speaker:about. For investors,
Speaker:the biggest thing that I would say is to,
Speaker:is in two areas. Number one, look for people that
Speaker:don't necessarily have the pedigree. It becomes really,
Speaker:really easy in quantum to assume
Speaker:that somebody must— that you got to have the PhD, and the more
Speaker:PhDs on the team, the better. And you see a lot of that.
Speaker:You look at D-Wave, Quantinuum, Qera, right? You
Speaker:look at all of these guys, and what you see is this long list of
Speaker:PhDs. And the truth is, is that companies like mine are
Speaker:completely blowing their doors off. Like, I don't—
Speaker:I'm rapidly getting to the point I don't even know how they're going to keep
Speaker:up. We have created a quantum
Speaker:error correction algorithm that
Speaker:reduces physical to logical overhead by 100. Wow.
Speaker:So yeah, it's better. It's better by so much.
Speaker:It's almost
Speaker:hard to believe. And we had to run it on
Speaker:the IBM Lima and Bellum benchmarks. We had to run
Speaker:that thing 3 times because we sort of assumed that
Speaker:it couldn't have been right the first time. You know, like, how
Speaker:is this good? So,
Speaker:so number one, look for people that are actually doing it.
Speaker:Number two, look for people that, that don't just need a check.
Speaker:Right, right. So for— too,
Speaker:too often investors are giving money to people either because
Speaker:of pedigree or because they go, oh, this guy has got,
Speaker:you know, two successful exits. So probably they can do a third one. But you
Speaker:look back at history and that's not true at all, right? Look at,
Speaker:look at Pets.com. Pets.com from, you know, way back in
Speaker:the dot-bomb era, right? It was started by multiple
Speaker:people that had done multiple different successful exits. And that
Speaker:thing was a disaster, right? No, it's true.
Speaker:And, you know, you mentioned that and 3DO, do you remember? Speaking
Speaker:of the '90s. Yes. 3DO, 3DO was like, I remember the
Speaker:Wired magazine article. Cover was like the digital start of the
Speaker:rise of the digital supergroup. And aside from like a handful of people
Speaker:who remember the '90s, no one knows what 3DL was, right?
Speaker:No, it was just like— and you're right, like pedigree. I think, I
Speaker:think there's a temptation. I think this brings up a deep point. Like, there's a
Speaker:temptation to overbuy
Speaker:on pedigree. Yes. Whether,
Speaker:whether it's in quantum, the assumption that, well, how could you possibly
Speaker:understand quantum if you don't have a PhD? And the answer is
Speaker:look at the solutions that are created, right? And then, and
Speaker:then the second thing, if somebody— just because somebody says I have
Speaker:something, or maybe they actually do— I mean, and this is something
Speaker:most venture capitalists or investors are already pretty good at,
Speaker:is going, let me see your customers. Um, you
Speaker:know, is there actually market traction for it?
Speaker:At the end of the day, a company— uh, so OA Quantum
Speaker:Labs isn't looking for an investor. But assuming that we were,
Speaker:we have— we don't only have solutions, we have customers. So because we have
Speaker:solutions and customers, like, I don't need your money. If I was going to—
Speaker:if I was going to take money from an investor,
Speaker:it would only be because that investor was bringing me
Speaker:some kind of strategic alliance that
Speaker:I like, that it's worth more than the equity that I
Speaker:would give up. Does that make sense? No, I mean, that makes sense. Yeah, no,
Speaker:I think, I think it's an interesting point you bring up. Like,
Speaker:um, it's about selling solutions. Yeah, not the science,
Speaker:right? It's almost like you— we got to give you a free copy of our
Speaker:book on, uh, selling quantum solutions, right?
Speaker:Um, because like, it, it's almost like you've read it. Like, because you're basically saying
Speaker:effectively the same thing, just like it. Yeah, you know that you're right.
Speaker:Like, if you can if you could prove the value— and I forget what it
Speaker:was, it was like months versus weeks— like, you could prove that real value to
Speaker:a business, doesn't really matter how many PhDs you have. I mean, obviously, right,
Speaker:obviously somebody has to, you know, check the numbers and make sure the answers you
Speaker:get are, you know, legit. Uh, but I mean, at some point
Speaker:it's really where the rubber meets the road, right? Like, I would not have thought,
Speaker:uh, if you look at pets.com compared to Amazon,
Speaker:um, Who, I mean, in the '90s, people would have assumed Pets.com would have won.
Speaker:Barnes Noble, like the same story. Fun fact, I worked at
Speaker:BarnesandNoble.com. Oh, wow. I was the first
Speaker:webmaster there.
Speaker:Wow. Underestimating people who are relentless is a
Speaker:mistake. Yes. Yeah. So when
Speaker:it comes to, let's say it's a developer who's interested
Speaker:in quantum, I would
Speaker:say that the best way for a developer to
Speaker:get involved in quantum is to get a Quantum Cloud account
Speaker:and to start creating
Speaker:stuff. Don't mess around with research.
Speaker:Don't, like, I mean, yeah, take some time to learn, you
Speaker:know, Qisk or whatever the
Speaker:different, DLLs that get wrapped up into Python.
Speaker:But yeah, take some time to learn what you're writing. But as soon as
Speaker:you can learn something, start creating something from
Speaker:it. Don't sit around and wait, create something from
Speaker:it because there is
Speaker:no substitute for experience. The
Speaker:problem that most developers have is that they've spent their entire
Speaker:lives either A, in school being taught, or
Speaker:B, in their careers on classical binary
Speaker:digital computers that are very time-dependent
Speaker:and sequentially processed. Whereas quantum
Speaker:is non-time-dependent and simultaneously processed.
Speaker:And therefore the way you have to even think
Speaker:about how you architect a
Speaker:solution is different. How you think about how you're going to write the code is
Speaker:different, and you don't know those things, or it's— I would
Speaker:be better to say it's hard to understand those things until
Speaker:you start writing the code and start seeing what
Speaker:happens, right? Right. So that's a good way to
Speaker:put it. Yeah. Sorry, Candice, I'll be quiet now. No, I'm just
Speaker:thinking about, you know, you come from such a
Speaker:unique background because most leaders are, you know, they're either in
Speaker:the quantum world or they're in the AI world.
Speaker:But because you're in both, it gives you such a
Speaker:unique advantage to have this dual
Speaker:fluency. So how do you find that that affects, you know, you
Speaker:being founder and CTO and CEO of of your
Speaker:company? It definitely in a lot of
Speaker:ways makes the commercial potential
Speaker:and applicability of what I'm
Speaker:doing better or easier. It means
Speaker:that when I am selling solutions, I
Speaker:can articulate to people like, this
Speaker:is, this is why what, what we're doing works
Speaker:better., right? And I'm able to
Speaker:speak to, you know, the CTOs of people. AI is
Speaker:getting to be understood well
Speaker:enough now in the enterprise and all those kinds
Speaker:of things that when I start to explain, okay, this is where
Speaker:the AI is and this is where the quantum is and this is why the
Speaker:quantum matters. It's a pretty simple conversation
Speaker:to have these days, especially now that they
Speaker:know that
Speaker:I'm bringing quantum enhancement. I'm not saying this
Speaker:is a quantum solution, it's a
Speaker:quantum enhancement. And that's— it's a very
Speaker:subtle distinction, but the gap between them is, you know,
Speaker:about the distance from one side of the Grand Canyon to
Speaker:the other. Well, it also frames the conversation differently. Sorry, Ken. No, I
Speaker:was thinking, so does that mean that the AI accelerates
Speaker:the quantum? No, the other way around. So the quantum accelerates
Speaker:the AI?
Speaker:Yes. Yes. Yeah. And, and it's because it's quantum
Speaker:accelerating the AI By having the discussion
Speaker:in that way, it means that
Speaker:the business people can understand it better. It means I
Speaker:can now have a much
Speaker:quicker conversation about this is what it means to your bottom line, because at the
Speaker:end of the day, that's what really matters, right?
Speaker:Right. If you're going to go to any enterprise, you had better be
Speaker:able to answer be able to say that either A,
Speaker:my solution is going to improve revenue, or B, it's going to
Speaker:reduce your cost and therefore improve profit. If you can't say
Speaker:it's going to do A, B, or both, don't
Speaker:even bother having the discussion because it
Speaker:doesn't matter, right? It's all about
Speaker:solutioning. Yes. Oh yes, not tech for the sake of tech. I mean, right, tech
Speaker:for the sake of tech is
Speaker:an academic conversation, correct? And that's fine for
Speaker:academia, but not outside of academia, correct? And
Speaker:I'm— and to
Speaker:me, one, quantum has gone far enough
Speaker:now that it no longer even should be
Speaker:in academia. And this is why you're seeing, even though everybody's— there's
Speaker:been a lot of news stories lately about, you know, the bursting of the
Speaker:quantum bubble. Or whatever. And D-Wave
Speaker:and Quantinuum and Rigetti have all been, let's call it punished a
Speaker:little bit. But the truth is, is that
Speaker:as we start having more of a
Speaker:business discussion, this is the business problems that we are solving
Speaker:right now today, the more that
Speaker:discussion goes away because now quantum starts moving into the data center
Speaker:and it really needs to get there for there to
Speaker:be additional significant investment investment to improve the technology.
Speaker:That makes a lot of sense.
Speaker:Yeah, yeah, we got to get it out of the research lab and into
Speaker:the enterprise. Very important.
Speaker:So when, if you're to look back earlier in,
Speaker:in your work, when, when you had
Speaker:that moment where you realized that this isn't just something theoretical,
Speaker:but this is actually something that I
Speaker:can commercialize What clicked
Speaker:for you? Okay, so one of the companies that
Speaker:I acquired is HughieBT. I was originally the CTO
Speaker:of HughieBT.
Speaker:Um, so
Speaker:at HughieBT, we have the most
Speaker:advanced digital identity solution by a very wide margin.
Speaker:Nobody else is even close. And, and it's a— we use
Speaker:behavioral biometrics as a way of, uh, it's—
Speaker:we are 99 point and then add
Speaker:7 nines percent of ability
Speaker:to distinguish between one human and another human.
Speaker:And because it's that accurate, it means we are that
Speaker:accurate also distinguishing between a deepfake. I have had
Speaker:people create deepfakes of themselves and not be able to defeat,
Speaker:um, our solution. Okay.
Speaker:Yeah, so the— when it clicked was
Speaker:when, um, the training was taking too long, and I was like, okay, well, what
Speaker:can I do? What can I do to this
Speaker:stupid thing? Um, and this is one of
Speaker:these weird
Speaker:sort of, um, so AI is an odd thing in general. It's, it
Speaker:can be odd. I, I am well known
Speaker:for saying that AI is
Speaker:really some shockingly simple
Speaker:algorithms, and it is about as intelligent actually
Speaker:as your calculator, right? Everybody wants to talk
Speaker:about, you know, is AI sentient? Is AI conscious? Is AI, you
Speaker:know, and how soon are we going to get to AGI? I don't think we're
Speaker:going to get to AGI anytime soon. I really don't.
Speaker:In fact, they've tried
Speaker:to change where AGI is from it being
Speaker:able to reason as good as a human to simply
Speaker:being able— being as— what's the word
Speaker:they use? Not learn. It's like adaptability or something. Like,
Speaker:they changed the bar for what's going to be
Speaker:considered AGI from reasoning capability to adaptability
Speaker:or something like that. And I just rolled my eyes and went, well, this
Speaker:is stupid. To me, it's not. Yeah, if you can't
Speaker:reason as good as even, you know,
Speaker:an average IQ person, then that's
Speaker:not artificial general intelligence. It's just not.
Speaker:So anyway, I Just on
Speaker:a lark, I asked the
Speaker:AI to consider itself as a
Speaker:high IQ materials scientist and to give
Speaker:me ways that I could
Speaker:improve the speed of training
Speaker:of the application. What it came up with
Speaker:was basically use quantum computing and it also
Speaker:output a whole bunch bunch of Cirq code. And I was
Speaker:like, well, this is interesting. So just as
Speaker:a hint to your audience, I know this is a
Speaker:quantum thing, this, but, but just as a sort of trick with
Speaker:AI, if you tell an AI
Speaker:to act in a role that
Speaker:is only sort of tertiary to
Speaker:what its actual thing that you're asking it to do, let's say you wanted to
Speaker:review code, Tell it that it's a chemist and to review
Speaker:the code from the viewpoint of a chemist. It
Speaker:will actually be more, for lack of a
Speaker:better word, creative. I know AI isn't actually
Speaker:creative, but sort of. It comes up with some
Speaker:really interesting responses that I have
Speaker:found dramatically improves it often,
Speaker:its output. Because of its having to, like
Speaker:I said, for lack of a better word, be creative. But anyway, so the
Speaker:very first thing that I did was
Speaker:implement quantum to improve the training
Speaker:of HuGPT. Then that grew into improving the
Speaker:inference of HuGPT. But in improving the
Speaker:inference of HuGPT, I sort of,
Speaker:by accident, for lack of a better way of putting it, created
Speaker:this system for how molecules and all of those kinds
Speaker:of things are modeled. I created a physics-informed
Speaker:neural network. Let me rephrase that, a
Speaker:quantum-enhanced physics-informed neural network. That then grew to where I
Speaker:was using PINs, PINOs. So,
Speaker:PIN, physics-informed neural network, physics-informed neural
Speaker:operator, GAN, which is a graph
Speaker:neural network, and a GNO, which is a graph
Speaker:neural operator. I started putting all of these things together and stuck quantum in the
Speaker:middle of it for doing the math, and then that
Speaker:grew into materials and grew into molecular modeling
Speaker:and all those
Speaker:things. It came to me, for lack of a
Speaker:better word, by
Speaker:accident because of output from, from
Speaker:an AI. And then just from deep
Speaker:diving into quantum is that's how these
Speaker:things happened.
Speaker:Interesting. What misconceptions do you run into the most when people
Speaker:hear AI plus quantum, and how do you try to
Speaker:reframe the conversation so they understand what's actually
Speaker:possible? The biggest one is they think I'm running the AI
Speaker:on quantum. Right? That's the biggest one. They go, you're running an AI on quantum.
Speaker:And then, you know, we go back into the whole, you know,
Speaker:is it conscious or whatever thing?
Speaker:And which I admittedly have a pretty
Speaker:low tolerance for. It irritates me when I hear,
Speaker:you know, people wanting to talk about you know, how intelligent
Speaker:they are, that they might be conscious or might be sentient or
Speaker:whatever. That stuff really is a pet peeve. I don't know why it drives me
Speaker:so crazy, but
Speaker:it does. But so anyway, that's the first misconception.
Speaker:The second misconception, and it comes from people
Speaker:within the AI industry, is the
Speaker:belief that quantum doesn't really
Speaker:have commercial application, that it doesn't really apply
Speaker:to AI, and Oh, you're just playing a game.
Speaker:You're not, you're not really doing what you're saying. You're not
Speaker:really doing blah, blah, blah. I'm like, you know, it's kind of hard to
Speaker:argue with the results, right? At the end of
Speaker:the day, you ask, you give me a problem domain for a
Speaker:material and I can spin out that material in 2 weeks. You tell me how
Speaker:I'm doing that without quantum enhancing a lot of
Speaker:different things. Right. And my nearest competitor needs 6
Speaker:months. The nearest competitor from them needs
Speaker:18 months. Right. Schrödinger needs
Speaker:18 months. Right. There's a lot to that, right? Like,
Speaker:you know, there's this idea of
Speaker:speed that Grant Cardone, one of the, one of my favorite kind of sales authors.
Speaker:Yeah. I love him. Yeah, everyone, you either love him or you hate him. There's
Speaker:nobody in the middle. But, um, you know, he has a phrase that, that really
Speaker:stuck with me. It's called speed is the new big.
Speaker:Yes, yes, 100% believe that. The phrase I use all the
Speaker:time is money loves
Speaker:speed. Yep. Oh, I like that. That's true too. That's
Speaker:quotable, right? I, I actually think I got that one from
Speaker:Jay Abraham. I don't know if you remember him or not, but he's another big
Speaker:sales guy from, from like the
Speaker:'90s. Interesting. So you're building in a field where things
Speaker:are evolving daily. Yeah. How do you
Speaker:stay ahead? Okay,
Speaker:so here's— things
Speaker:evolve fast. But when you're in the field,
Speaker:some of the times you almost wish they would
Speaker:evolve faster, especially in
Speaker:quantum. So quantum error
Speaker:correction has two separate problems. One, you want
Speaker:to maintain coherence for as long as possible, and number two, you
Speaker:want to prevent decoherence.
Speaker:Right? Two sides of the same fence, let's
Speaker:call it. So because of
Speaker:the nature of qubits and quantum and all of that kind of—
Speaker:and all of that, it's a lot
Speaker:harder. Those two pieces of the puzzle are a lot harder than it sounds. So
Speaker:even though I've got this really great
Speaker:quantum error correction algorithm where we can
Speaker:maintain coherence for about 4x longer, so instead
Speaker:instead of about
Speaker:300 microseconds, we're getting
Speaker:about 1.3 milliseconds we can maintain
Speaker:coherence, right? The best that we have been able to do on
Speaker:the decoherence side is predicting, well, these
Speaker:qubits are likely to decohere, therefore we can ignore those on the other side
Speaker:of the gate, right? 'Cause why
Speaker:pay attention? Sort of cut down on the amount of
Speaker:noise because we're ignoring the, we're
Speaker:ignoring the qubits that decohered. So you're almost doing quality
Speaker:assurance or QA on the qubits? Yeah, that's, that's actually a really good way
Speaker:of putting it.
Speaker:But we— there, there still needs to be a lot more work done
Speaker:in the lab on
Speaker:this aspect of preventing decoherence
Speaker:and maintaining coherence, because there's only so much that can
Speaker:be done on the software side, so let's call it, or
Speaker:the kernel side, where
Speaker:for that preventing decoherence
Speaker:or maintaining coherence, right? I can help
Speaker:the maintaining of coherence some, right? Like I said,
Speaker:extend it about 4x, But that's the best
Speaker:I can possibly get out of it. I'm not gonna get— 'cause now it's
Speaker:a hardware issue, right? I can only do
Speaker:so much. And this has been a problem for really a very,
Speaker:very, very long— since quantum started. And it
Speaker:hasn't, really hasn't improved a
Speaker:whole lot. So that's a
Speaker:big one. We still need a lot more work in the lab
Speaker:because until we can solve the coherence
Speaker:decoherence issue, scaling beyond about where we are
Speaker:now is going to be near impossible because there's just too
Speaker:much noise. What do you think it's going to take to solve
Speaker:that problem?
Speaker:Materials. Like new materials to be developed that the qubits, the quantum systems,
Speaker:are made out of?
Speaker:Absolutely. So I personally am
Speaker:convinced that that really is the
Speaker:major issue, is that part of the reason why we're having
Speaker:these coherence problems is that the materials
Speaker:aren't sufficient. And I can say I can spin
Speaker:out new materials once every 2 weeks, but number one, I can
Speaker:only sell so many. And getting these
Speaker:new materials through into the companies that are doing
Speaker:the research, IBM, Google,
Speaker:Continuum, Rigetti, QuEra, those guys,
Speaker:that can only happen so fast. Unless I'm physically part
Speaker:of their engineering teams, which of course I'm not. I've got my
Speaker:own company, right? So, um, you know, I would
Speaker:like to have a much more in-depth discussion with these
Speaker:guys about why, why their
Speaker:materials are causing the decoherence, even though I suspect they kind of
Speaker:know it, um, so that way those kinds of problems can be solved. But even
Speaker:once the new materials have been engineered, then they've gotta get actually
Speaker:into the QPU. Like, there's process
Speaker:that happens with this. So while from the outside,
Speaker:to go back to something, Candice, you had said before,
Speaker:it seems like things are
Speaker:moving fast, in a lot of ways, they still need to
Speaker:move faster. Because quantum, we
Speaker:need, AI right now is starting to bounce
Speaker:up against
Speaker:theoretical maximums. So because it's starting to bounce up against
Speaker:theoretical maximums, it's— this is why once we
Speaker:hit about GPT-3, you can almost
Speaker:draw a line there and you can see that the pace
Speaker:of AI improvement started slowing down and
Speaker:we started we stopped going from, it almost seemed
Speaker:like every few months there was
Speaker:this massive new improvements that we were getting out
Speaker:of AI. And lately all you're getting is
Speaker:incremental, very slow incremental
Speaker:improvements. Yeah, context windows are getting a little bit longer.
Speaker:Yeah, now it can remember past conversations a little bit better,
Speaker:or its ability to reason through code is slightly
Speaker:improved. But that's all we're getting. And we're getting
Speaker:that at the expense of massive
Speaker:new power requirements. Yeah, that's the big
Speaker:issue, isn't it? Yeah. Whereas if we
Speaker:can advance quantum a little bit faster, while
Speaker:quantum comes with a power requirement in the terms
Speaker:of cooling, the actual cost to run the QPU
Speaker:is almost zero. Right? It really doesn't cost
Speaker:a whole lot to run a QPU, right?
Speaker:Beyond the cooling, right? The cooling is, of course, is
Speaker:expensive, right? I'm not saying that the dilution fridges, you
Speaker:know, they're not cheap to buy them alone,
Speaker:you know, $500,000 per.
Speaker:And then the cost to run them is, you know, you got to keep it,
Speaker:you know, we're in dot Kelvin
Speaker:range, But assuming we can get beyond those
Speaker:with materials, we can find that we're
Speaker:able to push AI forward a
Speaker:lot because we really need to start running some of these
Speaker:neural networks, especially
Speaker:convolutional, on
Speaker:quantum. Right. Interesting. And for those that are not familiar with convolutional neural networks, they're
Speaker:a type of neural network architecture that's really good for
Speaker:image processing, typically. Images and video.
Speaker:Yeah. Yeah. Somebody had an experimental use case for
Speaker:them. I'm sorry, go ahead, Candace. No, I was wondering, has there
Speaker:been a breakthrough or a micro-innovation
Speaker:inside your lab that may not have made headlines
Speaker:but signals like a major shift in
Speaker:what's possible? Uh, well, there's been a few of them. Um, the fact that I
Speaker:can engineer new materials in 2 weeks, um, is a
Speaker:big one. Uh, that's, that's a really, really, really
Speaker:big one. Um, and, and like I said, we completely overran, um,
Speaker:the, you know, the ability of a sales team to even possibly
Speaker:keep up. Um, we have some, some of the new materials
Speaker:we have, um
Speaker:going to North American Stainless or US Steel,
Speaker:and then also to Ford. So some of the
Speaker:heat management, like heat shielding and all that kind of good stuff, Ford will
Speaker:be looking at shortly. But when you're talking about
Speaker:materials, when you create, engineer a material, especially when
Speaker:it's in silico, then they say, okay, well, now you got
Speaker:to create the thing. And then once you create the thing, then you have to
Speaker:test it, and then once you test it, then it has to get rolled into
Speaker:production and blah, blah, blah, right? Right. Yeah,
Speaker:but, but, um, so the materials is really, really a
Speaker:big one. Um, we haven't, um,
Speaker:uh, announced it broadly at all,
Speaker:uh, partially because it would be, it would be too
Speaker:easy to be too overwhelmed too fast.
Speaker:Um, so that's a big one. Um, the, the quantum error
Speaker:encryption would definitely be another one. We aren't talking—
Speaker:the only people we've talked to about our QEC so far
Speaker:is IBM. And that's because we benchmark it. We did our benchmarking
Speaker:on their systems. So, you know, they're a pretty easy one to have
Speaker:that discussion with. But, you know, IBM is a behemoth, right?
Speaker:It takes them— even IBM Quantum, it
Speaker:takes forever, you know, to get anywhere. But, you
Speaker:know, they're an obvious first place. And our quantum encryption algorithm
Speaker:is even agnostic. The, the— it doesn't matter whose hardware
Speaker:it is, it's going— the, the error correction is going to work no
Speaker:matter what, um, you know, whether it's D-Wave or Continuum
Speaker:or whoever. Um, so
Speaker:those would— yeah, the error correction would be a big one, and
Speaker:the, our— the materials, the— I, I say it's
Speaker:a quantum-enhanced pin, but that's only to simplify the discussion.
Speaker:There 4 different neural networks, and then what I call a
Speaker:controlling neural network that sits above it, and sort of in the middle, to think
Speaker:about it architecture-wise, in the middle is
Speaker:where the quantum enhancement sits, and the various neural
Speaker:networks sort of all talk to each other and talk
Speaker:to the quantum as a way of making this thing run
Speaker:so
Speaker:fast. Interesting. What— I know Candace usually asks
Speaker:this question. What's the biggest misconception out
Speaker:there about your business and kind of what you're up to? So
Speaker:there'll be— there, I guess I have to say there's two of them, and it
Speaker:depends on who you're talking to. Number one is in
Speaker:the business, too much of business outside of finance. This isn't so
Speaker:much true of— in the finance sector, they're starting to
Speaker:understand quantum. Because of QAOA, they understand, they're starting to understand. But outside
Speaker:of the financial vertical, there's still
Speaker:too much belief that quantum is a
Speaker:laboratory and research effort and there's no real
Speaker:true commercial applicability to it. And that's just
Speaker:completely false. That's probably the biggest
Speaker:one. And if we're talking
Speaker:about developers, I like to give this sort of as
Speaker:a hint. Most developers are
Speaker:using quantum computing almost like a
Speaker:trinary system, and I've had quantum developers disagree with me
Speaker:on this. And then I say, okay, let's look at your code, and I prove
Speaker:it to them. Most people
Speaker:are using quantum computers almost like a
Speaker:trinary computational device. You're gonna have to explain that. I roughly know what that
Speaker:is, but, uh, okay, so I like to be enlightened on terms of the
Speaker:difference between that because I've had this discussion and I didn't have a good
Speaker:answer to counter the statement. Yeah, okay, so, so
Speaker:right now, digital computers, classical computers, are binary,
Speaker:all right? It's a 1 or a 0 and that's it,
Speaker:okay? Most people are using, uh,
Speaker:quantum computers the same way. So think of 1 or 0 as spin up,
Speaker:spin down, right? And then you
Speaker:have superposition. Okay, so there— so trinary is spin up,
Speaker:spin down, superposition. Those are the
Speaker:3. Oh, okay. And they don't go any deeper with
Speaker:superposition. They stop there.
Speaker:Superposition actually means more than
Speaker:just both. Which is what superposition sort of means, but
Speaker:it doesn't just mean only
Speaker:that. Um,
Speaker:so it's— how do I put this in a way that I don't give away
Speaker:too much of my own secret sauce? Um,
Speaker:so you could treat— you could treat it that way. You could treat like a
Speaker:trinary system and you wouldn't be wrong, but you're not taking advantage
Speaker:of the superposition. Advantage, yes. Right. So for most normal
Speaker:people, I think a good way to look at this as,
Speaker:um, like a checkbox on a, on an online form, right? It's
Speaker:either checked, unchecked,
Speaker:or some designers will have a third space means you never touched
Speaker:it, right? So it's either kind of like yes, no, or unknown would be another
Speaker:one, right? Right. So you could almost use it like a maybe. So where the—
Speaker:what most developers are effectively doing it is using it like
Speaker:a maybe. Yes, no, maybe, on, off,
Speaker:don't know, right? But strictly speaking, superposition
Speaker:is yes and no at the same time. Correct.
Speaker:And, and it's not that it's yes, so yes, it's yes and no at the
Speaker:same time. But so let's, let's—
Speaker:I don't know, this might be getting a little bit deep into the woods, but
Speaker:let's look at Schrödinger's cat for a minute,
Speaker:right? Okay. The— in the thought experiment, it's the
Speaker:cat can be thought of as alive and dead at the same
Speaker:time. But here's the truth: the cat could also be
Speaker:thought of as in the process of
Speaker:dying. Okay, so it's not just alive and dead at the
Speaker:same time, it's alive, dead, and in the process of dying. And if it's in
Speaker:the process of dying, how far along the process
Speaker:of dying
Speaker:is it? Oh, okay,
Speaker:okay. So, so there's— there is a saying from,
Speaker:um, a personal development guy that I really like a lot.
Speaker:Most people in the personal development space, the minute they say
Speaker:quantum physics says, the next words that come out of their mouth
Speaker:are nonsense. Okay, nonsense. Uh,
Speaker:Dr. Joe Dispenza does a really good job when he says quantum
Speaker:physics says, the next words that come out of his mouth mouth are probably going
Speaker:to be right. When they are wrong,
Speaker:it's usually he's in the early part of
Speaker:his discussion, he's trying to get you to understand something, and if you listen to
Speaker:him a little bit longer, he makes it correct.
Speaker:So superposition is not just both. A better way
Speaker:of describing superposition would be it is a
Speaker:definition of all possible
Speaker:possibilities. Gotcha. Okay, that's what superposition actually is. So if
Speaker:it's all possible possibilities, that opens up more
Speaker:than just trinary computation. And that is about
Speaker:as far deep into the woods on that subject as I will
Speaker:get, because that's fair. We're at the top, we're towards the top of the hour,
Speaker:so like, it's, it's probably— we'd love to have you
Speaker:back because, uh, um, yeah, no, I, I, I feel you, like There's
Speaker:a lot to unpack there. I can kind of sense like,
Speaker:oh, this— it's kind of like you pull a thread on a sweater, like, oh,
Speaker:it's not just this little bit. It's actually way more
Speaker:than I anticipated. Yeah. And then you start getting into the
Speaker:woods of what is superposition exactly and whether it's,
Speaker:you know, we're talking about when you're measuring spin up,
Speaker:spin down, it's because you have collapsed the particle
Speaker:into a into a particle, and superposition is actually just
Speaker:the waveform, right? And that's right. And then you start getting the double slit and
Speaker:all those kinds of fun things. Yeah,
Speaker:causality, you start— yeah, yeah, yeah. So anyway, I can,
Speaker:I can geek out on this for kind of— we'd love to have another— you
Speaker:back on the show. Yeah, absolutely. This has been a fantastic
Speaker:conversation. We really, really appreciate your time. I thank you. I've had a lot
Speaker:of fun. They're skanking, skanking in time.
Speaker:Black holes are wailing in a horn line so fine. From plank scales to
Speaker:planets, they're connecting the dots. Candace and Frank, they're
Speaker:the
Speaker:cosmic hotshot. Quantum Podcast, turn it up fast. Candace
Speaker:and Frank blowing my mind at last. Quantum Podcast,
Speaker:they're breaking the mold. Science has got beats.
Speaker:It's bold and it's gold.
Leave a Reply